Nvidia Launches 64GB DGX Spark for Local AI Development
Nvidia has unveiled a new 64GB version of its DGX Spark personal AI supercomputer, marking another step in its efforts to enable developers to run large AI models locally rather than relying on cloud-based solutions.
The device will be available starting October 23 through partners such as Acer, ASUS, Dell, Gigabyte, HP, and MSI, with a starting price of $4,999. It features the same GB10 Grace Blackwell Superchip and Nvidia AI software stack as the existing 128GB version.
According to Nvidia, the 64GB system can handle models with up to 100 billion parameters on-device, catering to developers working on AI agents, inference, fine-tuning, and other local AI tasks.
The new system also supports clustering, allowing two 64GB DGX Spark systems to be connected for a combined 128GB of memory, which can support models with up to 200 billion parameters. Nvidia reported that internal testing with the Qwen 3.8 27B model showed performance gains of up to 1.7 times compared to a single system.